Here, a water quality assessment model (WQAM) is developed by non-linear regression as an alternative to physical watershed modeling in South Korea. Three cases and 10 scenarios are applied and reviewed to determine the most appropriate WQAM. The three cases are: (1) the area size allocation of sub-watersheds, (2) the watershed imperviousness ratio, and (3) the combination of the area size and imperviousness ratio. The 10 scenarios are: (1) impervious, (2) impervious + pervious, (3) impervious + rainfall, (4) impervious + slope, (5) impervious + rainfall + slope, (6) slope, (7) land usage, (8) land usage + rainfall, (9) land usage + slope, and (10) land usage + rainfall + slope. The best WQAMs are subsequently developed from the generated equations using statistics (R2, Adjusted R2, F-test, the Akaike information criterion and the Shapiro–Wilk test). In addition, the WQAM is verified using the Geum-Sum-Young River watershed. The percentage differences of biochemical oxygen demand (BOD), total nitrogen (T-N), and total phosphorus (T-P) are 31.66%, 8.08%, and 48.94%, respectively. The developed WQAM can be used in place of complex watershed modeling and to aid in the determination of the best restoration locations.

INTRODUCTION

Rapid population growth, land-use conversion and its accompanying pollution are major problems today (Novotny 2003). Due to recent trends in land use, imperviousness has increased abruptly. These changes have affected the hydrographic conditions of urban streams and have resulted in significantly higher and earlier peak discharge rates as compared to streams in rural or undeveloped areas (Schueler 1987, 2005). In order to prioritize rural and urban areas for costly restoration or rehabilitation projects, policy makers will need to evaluate the relationship between the land coverage and the water quality for each area. To implement these evaluations, we need to adapt nationwide watershed modeling for forecasting after applying watershed restoration measures (Singh & Frevert 2004; Novotny 2008). However, existing watershed modeling methodologies are extremely complicated and time-consuming. Therefore, simplified assessment methods for watershed environmental conditions should be developed to forecast the effect of the restoration. The objective of this research is to develop efficient water quality assessment models and develop a decision-making system that utilizes existing data, such as land coverage and watershed area data (Figure 1).

Figure 1

Research schematic for developing a WQAM.

Figure 1

Research schematic for developing a WQAM.

DATA COLLECTION FOR DEVELOPING WATER QUALITY ASSESSMENT MODEL (WQAM)

Numerous studies have been published on the impact correlation between land use types and water quality parameters (Sliva & Williams 2001; Mehaffey et al. 2005; Schoonover et al. 2005; Woli et al. 2004; Stutter et al. 2007; Tu 2011). In order to determine the relationship between water quality and these parameters, the data shown in Table 1 were collected.

Table 1

The data sources and periods in order to compare land usage and water quality

DivisionParameterSourcePeriod
Water Quality COD, BOD, T-N, T-P MOE 10 years (2001–2010) 
Hydrology Rainfall WAMIS 30 years (1966–2007) 
Geology Slope   
Land Usage Pervious & impervious, urban, agriculture, forest, grass, wetland, barren, water Remote sensing. Landsat TM data were collected between 2008 and 2010 at a 30*30 m spatial resolution and were processed in order to reveal the land use/land change features. Geometrical corrections, classification and accuracy assessments of the data were carried out with the support of PCI Geomatics digital image processing software. Topographical maps (1:50,000) were used as the reference for the geometric corrections. According to the geographical names, the maps were re-drawn on the digitized topographical map base using ArcView GIS 3.2 ver. software to complete the editing, labeling, projection, transformation, edge matching and overlaying processes. 2008–2010 
DivisionParameterSourcePeriod
Water Quality COD, BOD, T-N, T-P MOE 10 years (2001–2010) 
Hydrology Rainfall WAMIS 30 years (1966–2007) 
Geology Slope   
Land Usage Pervious & impervious, urban, agriculture, forest, grass, wetland, barren, water Remote sensing. Landsat TM data were collected between 2008 and 2010 at a 30*30 m spatial resolution and were processed in order to reveal the land use/land change features. Geometrical corrections, classification and accuracy assessments of the data were carried out with the support of PCI Geomatics digital image processing software. Topographical maps (1:50,000) were used as the reference for the geometric corrections. According to the geographical names, the maps were re-drawn on the digitized topographical map base using ArcView GIS 3.2 ver. software to complete the editing, labeling, projection, transformation, edge matching and overlaying processes. 2008–2010 

Water quality data were obtained from the Ministry of Environment (MOE) from 2001 to 2010. For rainfall, the 30-year average rainfall was used per standard basin from 1966 to 2007, as obtained from the Water Management Information System (WAMIS). Land usage data are derived from land cover maps (scale of 1:50,000) which were photographed by LANSAT 7 from 2008 to 2010. This was published by the MOE, Republic of Korea, in 2010.

PARAMETER SELECTION PROCESS

Tu (2011) researched the relationship between six land usage types (agricultural, forest, commercial, industrial, residential, and recreation land) and fourteen water quality indicators in Eastern Massachusetts. Most of the water quality indicators were significantly associated with most of the land use indicators. Randhir et al. (2001) studied a watershed-based land prioritization model for water supply protection based on the integration of three types of information: geographic information, the relationships between land criteria and effects, and travel-time of runoff. In land use modeling studies, geographic information systems (GIS) have been used for assembling data and defining decision zones (Wang et al. 2004).

In this research, in order to develop the WQAM, we assumed three cases. The first case takes into consideration the area size of sub-watersheds, the second case considers the imperviousness ratio, and the last case considers both the area size and the imperviousness ratio.

Pertaining to the area size allocation of the sub-watershed, the Korean peninsula was divided into the watersheds of three rivers. The co-relationships between land use and water quality are not constant in different regions because the characteristics and pollution sources of watersheds are not the same in different places (Tu 2011). The three groups of watersheds are the Han River, the Nakdong River, and the Geum-Sum-Youngsan River, as shown in Figure 2. The first case, involving the correlation of water quality, hydrology, geology, and land usage, is to classify the sub-watersheds into bins according to size of the area: 0–200, 200–500, and over 500 km2, for the Han and Nakdong River watersheds, and 0–100, 100–150, 150–200 km2, and over 200 km2 for the Geum-Sum-Youngsan River watershed, as shown in Table 2. In the second case, watershed data are classified in accordance with imperviousness, because this factor is closely related to the water-quality impact (Conway 2007). In this case, imperviousness is divided into four intervals: below 20%, 20–25%, and over 30% as shown in Table 3. In order to consider both sub-watershed areas and the percentage of imperviousness among the sub-watershed characteristics, sub-watershed areas are divided below and above 250 km2 and their imperviousness cover is broken up into categories of 0–20%, 20–25%, and over 25%, as shown Table 4.

Table 2

The number of applied sub-watersheds for each river watershed (case one)

Number of applied sub-watersheds
Area (km2)Han RiverNakdong RiverGeum-Sum-Youngsan River
0 ∼ 100 – – 14 
100 ∼ 150 – – 17 
150 ∼ 200 – – 13 
200 ∼ – – 38 
0 ∼ 200 22 17 – 
200 ∼ 500 23 28 – 
500 ∼ 16 19 – 
Total (207) 61 64 82 
Number of applied sub-watersheds
Area (km2)Han RiverNakdong RiverGeum-Sum-Youngsan River
0 ∼ 100 – – 14 
100 ∼ 150 – – 17 
150 ∼ 200 – – 13 
200 ∼ – – 38 
0 ∼ 200 22 17 – 
200 ∼ 500 23 28 – 
500 ∼ 16 19 – 
Total (207) 61 64 82 
Table 3

The number of applied sub-watersheds for each river watershed (case two)

The number of sub-watersheds
ImperviousnessTotalHan RiverNakdong RiverGeum-Sum-Youngsan River
∼ 20% 75 33 21 21 
20 ∼ 25% 88 19 35 34 
25% ∼ 36 19 
The number of sub-watersheds
ImperviousnessTotalHan RiverNakdong RiverGeum-Sum-Youngsan River
∼ 20% 75 33 21 21 
20 ∼ 25% 88 19 35 34 
25% ∼ 36 19 
Table 4

The number of standard basins for both the area allocation and the percentage of imperviousness of the standard basins (case three)

Imperviousness (%)
Area allocation (km2)Below 20%20 ∼ 25%Over 25%
Below 250 km2 23 53 24 
Over 250 km2 53 41 13 
Imperviousness (%)
Area allocation (km2)Below 20%20 ∼ 25%Over 25%
Below 250 km2 23 53 24 
Over 250 km2 53 41 13 
Figure 2

Watershed map of the three groups of South Korea.

Figure 2

Watershed map of the three groups of South Korea.

To identify the critical parameters that have strong relationships with water quality, five parameters – impervious, pervious, rainfall, slope, and land usage – are combined with imperviousness and land usage in the following 10 scenarios as shown in Table 5.

Table 5

Data classification and scenarios for generating WQAM

CasesWatershedRangekm and percentage of land use (Attribute)Scenarios*Total scenarios
First (area size) Han river watershed Below 200 km2 (01) km (land use), (02) Percentage (land use) (01) imp, (02) imp, per, (03) imp, ra, (04) imp, ra, sl, (05) imp, sl, (06) sl, (07) land, (08) land, ra, (09) land, sl, (10) land, ra, sl 60 (han) 60 (Nak) 80 (GSY) Total: 200 
200 ∼ 500 km2 
Over 500 km2 
Nakdong river watershed Below 200 km2 
200 ∼ 500 km2 
Over 500 km2 
Geum-Sum-Youngsum river watershed Below 100 km2 
100 ∼ 150 km2 
150 ∼ 200 km2 
Over 200 km2 
Second (impervious ratio) Han river watershed Below 20% (01) km (land use), (02) Percentage (land use) (01) imp, (02) imp, per, (03) imp, ra, (04) imp, ra, sl, (05) imp, sl, (06) sl, (07) land, (08) land, ra, (09) land, sl, (10) land, ra, sl 60 (han) 60 (Nak) 60 (GSY) Total: 180 
20 ∼ 25% 
Over 25% 
Nakdong river watershed Below 20% 
20 ∼ 25% 
Over 25% 
 Geum-Sum-Youngsum river watershed Below 20% 
20 ∼ 25% 
Over 25% 
Third (area size + Impervious ratio) Below 250 km2 Below 20% (01) km (land use), (02) Percentage (land use) (01) imp, (02) imp, per, (03) imp, ra, (04) imp, ra, sl, (05) imp, sl, (06) sl, (07) land, (08) land, ra, (09) land, sl, (10) land, ra, sl 60 (be250) 60 (ov250) Total:120 
20 ∼ 25% 
Over 25% 
Over 250 km2 Below 20% 
20 ∼ 25% 
Over 25% 
CasesWatershedRangekm and percentage of land use (Attribute)Scenarios*Total scenarios
First (area size) Han river watershed Below 200 km2 (01) km (land use), (02) Percentage (land use) (01) imp, (02) imp, per, (03) imp, ra, (04) imp, ra, sl, (05) imp, sl, (06) sl, (07) land, (08) land, ra, (09) land, sl, (10) land, ra, sl 60 (han) 60 (Nak) 80 (GSY) Total: 200 
200 ∼ 500 km2 
Over 500 km2 
Nakdong river watershed Below 200 km2 
200 ∼ 500 km2 
Over 500 km2 
Geum-Sum-Youngsum river watershed Below 100 km2 
100 ∼ 150 km2 
150 ∼ 200 km2 
Over 200 km2 
Second (impervious ratio) Han river watershed Below 20% (01) km (land use), (02) Percentage (land use) (01) imp, (02) imp, per, (03) imp, ra, (04) imp, ra, sl, (05) imp, sl, (06) sl, (07) land, (08) land, ra, (09) land, sl, (10) land, ra, sl 60 (han) 60 (Nak) 60 (GSY) Total: 180 
20 ∼ 25% 
Over 25% 
Nakdong river watershed Below 20% 
20 ∼ 25% 
Over 25% 
 Geum-Sum-Youngsum river watershed Below 20% 
20 ∼ 25% 
Over 25% 
Third (area size + Impervious ratio) Below 250 km2 Below 20% (01) km (land use), (02) Percentage (land use) (01) imp, (02) imp, per, (03) imp, ra, (04) imp, ra, sl, (05) imp, sl, (06) sl, (07) land, (08) land, ra, (09) land, sl, (10) land, ra, sl 60 (be250) 60 (ov250) Total:120 
20 ∼ 25% 
Over 25% 
Over 250 km2 Below 20% 
20 ∼ 25% 
Over 25% 

*imp: impervious, per: pervious, ra: rain, sl: slope, land: land usage.

Land usage: urban, agriculture, forest, grass, wetland, barren, water (seven items).

METHODS AND ARCHITECTURE

The WQAMs are established through the appliation of the three cases and 10 scenarios in a non-linear regression method, as shown in Figure 3. In order to identify the best WQAM, two processes are reviewed; a model selection process and an evaluation/validation model with parameter estimation (Kutner et al. 2004).

Figure 3

Architecture to establish the WQAM based on the data analysis methods.

Figure 3

Architecture to establish the WQAM based on the data analysis methods.

Non-linear regression analysis

Three cases are compared in order to select the best cases based on their R2 values. Subsequently, five steps are undertaken, as follows: formulation of the WQAM based on the 10 scenarios (first step), minimization of the difference between the observed and predicted water quality concentration based on the 10 scenarios (second step), calculation of the coefficient of multiple determination (R2), and adjusted coefficient of multiple determination (R2) to compare the efficiency of the water quality model, execution of an F-test to test the overall significance of the water quality equations (a P-value less than 0.05 was deemed significant), evaluation of the trade-off between bias and variance using the Akaike information criterion (AIC), determination of the relative goodness of fit, and selection of the most appropriate WQAM (third step). The best WQAM is then selected based on the model which has the AIC smallest value (fourth and fifth step) and the WQAMs are tested with the Shapiro–Wilk test for model evaluation with parameter analyses (Pruden et al. 2012). Examples of scenarios seven and the abovementioned five steps are shown in Figure 4.

Figure 4

Procedure to select the best WQAM based on statistics' evaluation.

Figure 4

Procedure to select the best WQAM based on statistics' evaluation.

RESULTS

The best WQAMs are derived from the first, second, and third cases by means of non-linear regression. The R2 values of the WQAMs in the first cases are higher than those of the second and third cases, except for that of the Nakdong River, as shown in Figure 5 and Table 6. Hence, the first WQAM is selected as the best case.

Table 6

A detailed explanation of the range for the index shown in Figure 5 

CasesWatershedRangeIndexCasesWatershedRangeIndex
First step (area size) Han river watershed Below 200 km2 FH0 ∼ 200 Second (imper- vious ratio) Nakdong river watershed Below 20% SN0 ∼ 20% 
200 ∼ 500 km2 FH200 ∼ 500 20 ∼ 25% SN20 ∼ 25% 
Over 500 km2 FH500 ∼ Over 25% SN25% ∼ 
Nakdong river watershed Below 200 km2 FN0 ∼ 200 Geum-sum-youngsum river watershed Below 20% SG0 ∼ 20% 
200 ∼ 500 km2 FN200 ∼ 500 20 ∼ 25% SG20 ∼ 25% 
Over 500 km2 FN500 ∼ Over 25% SG25% ∼ 
Geum-sum-youngsum river watershed Below 100 km2 FG0 ∼ 100 Third (area size + impervious ratio) Below 250 Km2 Below 20% TB250_IP0 ∼ 20% 
100 ∼ 150 km2 FG100 ∼ 150 20 ∼ 25% TB250_IP20 ∼ 25% 
150 ∼ 200 km2 FG150 ∼ 200 Over 25% TB250_IP25% ∼ 
Over 200 km2 FG200 ∼ Over 250 Km2 Below 20% TO250_IP0 ∼ 20% 
Second (imper- vious ratio) Han river watershed Below 20% SH0 ∼ 20% 20 ∼ 25% TO250_IP20 ∼ 25% 
20 ∼ 25% SH20 ∼ 25% Over 25% TO250_IP25% ∼ 
Over 25% SH25% ∼         
CasesWatershedRangeIndexCasesWatershedRangeIndex
First step (area size) Han river watershed Below 200 km2 FH0 ∼ 200 Second (imper- vious ratio) Nakdong river watershed Below 20% SN0 ∼ 20% 
200 ∼ 500 km2 FH200 ∼ 500 20 ∼ 25% SN20 ∼ 25% 
Over 500 km2 FH500 ∼ Over 25% SN25% ∼ 
Nakdong river watershed Below 200 km2 FN0 ∼ 200 Geum-sum-youngsum river watershed Below 20% SG0 ∼ 20% 
200 ∼ 500 km2 FN200 ∼ 500 20 ∼ 25% SG20 ∼ 25% 
Over 500 km2 FN500 ∼ Over 25% SG25% ∼ 
Geum-sum-youngsum river watershed Below 100 km2 FG0 ∼ 100 Third (area size + impervious ratio) Below 250 Km2 Below 20% TB250_IP0 ∼ 20% 
100 ∼ 150 km2 FG100 ∼ 150 20 ∼ 25% TB250_IP20 ∼ 25% 
150 ∼ 200 km2 FG150 ∼ 200 Over 25% TB250_IP25% ∼ 
Over 200 km2 FG200 ∼ Over 250 Km2 Below 20% TO250_IP0 ∼ 20% 
Second (imper- vious ratio) Han river watershed Below 20% SH0 ∼ 20% 20 ∼ 25% TO250_IP20 ∼ 25% 
20 ∼ 25% SH20 ∼ 25% Over 25% TO250_IP25% ∼ 
Over 25% SH25% ∼         
Figure 5

Comparing R2 values among first, second, and third WQAMs based on non-linear regression.

Figure 5

Comparing R2 values among first, second, and third WQAMs based on non-linear regression.

Table 7 shows the best WQAM of biochemical oxygen demand (BOD) simulation derived from the first cases. It should be noted that water quality is significantly related to watershed land usage, rainfall, slope, imperviousness, and perviousness. In the BOD simulation, Geum-Sum-Youngsan River from 150 km2 to 200 km2 gives the best WQAM [F = 4,296, P < 0.001, df = 9, R2 = 0.998]. The Nakdong River from 0 km2 to 200 km2 gives the worst WQAM [F = 9, P = 0.008, df = 19, R2 = 0.319]. The results of the chemical oxygen demand (COD), T-N, and T-P simulations are shown in Supplementary Information Table S1 (available online at http://www.iwaponline.com/ws/015/098.pdf).

Table 7

The WQAM for the water quality simulation based on non-linear regression

Simple equationNormality (Shapiro)Selection
RiverParametersScenarioArea (km2)WQAM*R2Adj R2Fp-valuend.fWp-valueSSE(Akaike) AICAttribute
Han river Land usage/rainfall/slope 10 0–200 BOD(mg/L) = 2.66 Ur1.30 Ag0.29 Fo0.42 Gr−0.54 Wet0.03 Ba−0.41 Wa−0.22 Ra−0.68 SI−0.61 0.974 0.973 762 <2.2e-16 22 20 0.671 0.000 3.42 −22.95 km2 
 Land usage/rainfall 200–500 BOD(mg/L) = 2.33 Ur0.47 Ag−0.51 Fo−0.64 Gr0.20 Wet0.26 Ba0.35 Wa−0.64 Ra−0.90 0.988 0.988 1,788 <2.2e-16 23 21 0.604 0.000 5.10 −18.64 km2 
 Land usage/slope 500– BOD(mg/L) = 1.53 Ur0.12 Ag–0.13 Fo1.83 Gr0.27 Wet−0.47 Ba0.54 Wa−0.04 SI−2.40 0.962 0.959 324 0.000 15 13 0.676 0.000 0.44 −37.04 
Nakdong river Impervious 0–200 BOD(mg/L) = 0.01 IP1.97 0.319 0.284 0.008 21 19 0.888 0.021 13.03 −8.03 
 Impervious 200–500 BOD(mg/L) = 0.01 IP2.16 0.335 0.313 15 0.001 32 30 0.893 0.004 13.98 −24.50 
 Land usage/rainfall 500– BOD(mg/L) = 1.91 Ur0.39 Ag−0.06 Fo−0.65 Gr0.45 Wet−0.07 Ba−0.61 Wa0.28 Ra0.47 0.985 0.983 573 0.000 11 0.775 0.004 0.09 −36.50 
Geun-Sum-Youngsan river Land usage/rainfall/slope 10 0–100 BOD(mg/L) = 3.32 Ur0.48 Ag−0.98 Fo2.04 Gr−0.69 Wet0.22 Ba−0.40 Wa0.09 Ra0.27 SI−2.24 0.863 0.851 75 0.000 14 12 0.856 0.027 1.26 −15.70 km2 
 Land usage/slope 100–150 BOD(mg/L) = 1.95 Ur0.00 Ag0.33 Fo2.62 Gr0.21 Wet−0.51 Ba0.00 Wa−0.28 SI−3.63 0.919 0.912 136 0.000 14 12 0.886 0.071 2.61 −7.51 
 Land usage/rainfall 150–200 BOD(mg/L) = 2.85 Ur1.16 Ag−0.40 Fo1.54 Gr0.10 Wet−0.55 Ba−0.71 Wa1.63 Ra−1.99 0.998 0.998 4,296 0.000 11 0.882 0.111 0.02 −52.36 km2 
 Pervious/impervious 200– BOD(mg/L) = 3.45 P−1.98 IP2.38 0.643 0.632 61 0.000 36 34 0.832 0.000 21.16 −15.13 km2 
Simple equationNormality (Shapiro)Selection
RiverParametersScenarioArea (km2)WQAM*R2Adj R2Fp-valuend.fWp-valueSSE(Akaike) AICAttribute
Han river Land usage/rainfall/slope 10 0–200 BOD(mg/L) = 2.66 Ur1.30 Ag0.29 Fo0.42 Gr−0.54 Wet0.03 Ba−0.41 Wa−0.22 Ra−0.68 SI−0.61 0.974 0.973 762 <2.2e-16 22 20 0.671 0.000 3.42 −22.95 km2 
 Land usage/rainfall 200–500 BOD(mg/L) = 2.33 Ur0.47 Ag−0.51 Fo−0.64 Gr0.20 Wet0.26 Ba0.35 Wa−0.64 Ra−0.90 0.988 0.988 1,788 <2.2e-16 23 21 0.604 0.000 5.10 −18.64 km2 
 Land usage/slope 500– BOD(mg/L) = 1.53 Ur0.12 Ag–0.13 Fo1.83 Gr0.27 Wet−0.47 Ba0.54 Wa−0.04 SI−2.40 0.962 0.959 324 0.000 15 13 0.676 0.000 0.44 −37.04 
Nakdong river Impervious 0–200 BOD(mg/L) = 0.01 IP1.97 0.319 0.284 0.008 21 19 0.888 0.021 13.03 −8.03 
 Impervious 200–500 BOD(mg/L) = 0.01 IP2.16 0.335 0.313 15 0.001 32 30 0.893 0.004 13.98 −24.50 
 Land usage/rainfall 500– BOD(mg/L) = 1.91 Ur0.39 Ag−0.06 Fo−0.65 Gr0.45 Wet−0.07 Ba−0.61 Wa0.28 Ra0.47 0.985 0.983 573 0.000 11 0.775 0.004 0.09 −36.50 
Geun-Sum-Youngsan river Land usage/rainfall/slope 10 0–100 BOD(mg/L) = 3.32 Ur0.48 Ag−0.98 Fo2.04 Gr−0.69 Wet0.22 Ba−0.40 Wa0.09 Ra0.27 SI−2.24 0.863 0.851 75 0.000 14 12 0.856 0.027 1.26 −15.70 km2 
 Land usage/slope 100–150 BOD(mg/L) = 1.95 Ur0.00 Ag0.33 Fo2.62 Gr0.21 Wet−0.51 Ba0.00 Wa−0.28 SI−3.63 0.919 0.912 136 0.000 14 12 0.886 0.071 2.61 −7.51 
 Land usage/rainfall 150–200 BOD(mg/L) = 2.85 Ur1.16 Ag−0.40 Fo1.54 Gr0.10 Wet−0.55 Ba−0.71 Wa1.63 Ra−1.99 0.998 0.998 4,296 0.000 11 0.882 0.111 0.02 −52.36 km2 
 Pervious/impervious 200– BOD(mg/L) = 3.45 P−1.98 IP2.38 0.643 0.632 61 0.000 36 34 0.832 0.000 21.16 −15.13 km2 

*IP: impervious, P: pervious, Ra: rain, Sl: slope, land: land usage.

Ur: urban, Ag: agriculture, Fo: forest, Gr: grass, Wet: wetland, Ba: barren, Wa: water (7 items of land usage).

Attribute: km2: impervious, pervious, and land usage were calculated by area size (km2).

%: impervious, pervious, and land usage were calculated by percentage (%) of area.

VERIFICATION

WQAMs are verified in order to demonstrate their reliability. The Yongdam Dam watershed, which is located at 127°18′49″ to 127°44′47″ east longitude and 35°34′50″to 36°1′37″ north latitude in the southern part of Korea, is used for the verification exercise in order to simulate the water quality; the WQAM shown in Table 7 and Supplementary Information Table S1 are used.

The data sets for simulations of the Yongdam Dam watershed are shown in Table 8. They include the pervious, impervious, rainfall, slope, and land usage data sets. Land usage is from Landsat TM data (Table 1). Water quality data are adapted from the end site of each sub-watershed. In particular, the average of 2010 water quality is adapted.

Table 8

Data sets for the water quality simulation using the WQAM (Geum-Sum-Youngsan river watershed)

Land useWater quality (average from 2001 to 2010)
WatershedSub-watershedPerviousImperviousRainfall (mm/month)Slope (%)UrbanAgricultureForestGrassWetlandBarrenWaterTotalBOD (mg/L)T-N(mg/L)T-P(mg/L)
Geum-Sum-Youngsan river Watershed Geum River (km2224 57.70 112.4 32.87 8.43 74.52 178.14 16.61 1.03 3.42 2.24 282.157 2.12 2.717 0.151 
 Geum River (%) 80 20 112.4 32.87 2.99 26.41 63.13 5.89 0.37 1.21 0.79 100 
 Guryang Stream (km2129 33.23 146.9 29.75 4.59 44.67 109.23 2.59 0.21 1.33 0.98 162.629 2.45 2.683 0.158 
 Guryang Stream (%) 80 20 146.9 29.75 2.82 27.47 67.17 1.59 0.13 0.82 0.60 100 
 Jinan Stream (km227 7.63 113.7 26.98 1.99 10.74 20.39 1.12 0.16 0.13 0.20 34.517 1.93 1.862 0.096 
 Jinan Stream (%) 78 22 113.7 26.98 5.77 31.11 59.07 3.24 0.46 0.36 0.57 100 
 Jeongia Stream (km280 16.68 134.4 40.46 1.34 11.71 81.95 1.64 0.15 0.17 0.48 96.954 1.29 2.577 0.026 
 Jeongia Stream (%) 83 17 134.4 40.46 1.38 12.08 84.52 1.69 0.15 0.18 0.49 100 
 Juja Stream (km248 9.40 137.8 40.23 0.62 5.21 50.22 0.74 0.10 0.16 0.35 57.038 1.17 1.615 0.015 
 Juja Stream (%) 84 16 137.8 40.23 1.08 9.14 88.04 1.29 0.17 0.28 0.61 100 
Land useWater quality (average from 2001 to 2010)
WatershedSub-watershedPerviousImperviousRainfall (mm/month)Slope (%)UrbanAgricultureForestGrassWetlandBarrenWaterTotalBOD (mg/L)T-N(mg/L)T-P(mg/L)
Geum-Sum-Youngsan river Watershed Geum River (km2224 57.70 112.4 32.87 8.43 74.52 178.14 16.61 1.03 3.42 2.24 282.157 2.12 2.717 0.151 
 Geum River (%) 80 20 112.4 32.87 2.99 26.41 63.13 5.89 0.37 1.21 0.79 100 
 Guryang Stream (km2129 33.23 146.9 29.75 4.59 44.67 109.23 2.59 0.21 1.33 0.98 162.629 2.45 2.683 0.158 
 Guryang Stream (%) 80 20 146.9 29.75 2.82 27.47 67.17 1.59 0.13 0.82 0.60 100 
 Jinan Stream (km227 7.63 113.7 26.98 1.99 10.74 20.39 1.12 0.16 0.13 0.20 34.517 1.93 1.862 0.096 
 Jinan Stream (%) 78 22 113.7 26.98 5.77 31.11 59.07 3.24 0.46 0.36 0.57 100 
 Jeongia Stream (km280 16.68 134.4 40.46 1.34 11.71 81.95 1.64 0.15 0.17 0.48 96.954 1.29 2.577 0.026 
 Jeongia Stream (%) 83 17 134.4 40.46 1.38 12.08 84.52 1.69 0.15 0.18 0.49 100 
 Juja Stream (km248 9.40 137.8 40.23 0.62 5.21 50.22 0.74 0.10 0.16 0.35 57.038 1.17 1.615 0.015 
 Juja Stream (%) 84 16 137.8 40.23 1.08 9.14 88.04 1.29 0.17 0.28 0.61 100 

The results of water quality simulation using the WQAM are shown in Table 9 and Figure 6.

Table 9

The results of the water quality simulation for Yongdam Dam watershed using the WQAM

BOD (mg/L)T-N (mg/L)T-P (mg/L)
Sub-watershedWQAMObservedWQAMObservedWQAMObserved
Geum river 0.80 2.12 1.44 2.72 0.039 0.151 
Guryang stream 0.47 2.45 2.71 2.68 0.077 0.158 
Jinan stream 0.57 1.93 2.02 1.86 0.010 0.096 
Jeongja stream 2.25 1.29 2.38 2.58 0.068 0.026 
Juja stream 2.03 1.17 1.98 1.62 0.034 0.015 
sum 6.13 8.97 10.53 11.45 0.227 0.446 
% difference 31.66 – 8.08 – 48.948 – 
BOD (mg/L)T-N (mg/L)T-P (mg/L)
Sub-watershedWQAMObservedWQAMObservedWQAMObserved
Geum river 0.80 2.12 1.44 2.72 0.039 0.151 
Guryang stream 0.47 2.45 2.71 2.68 0.077 0.158 
Jinan stream 0.57 1.93 2.02 1.86 0.010 0.096 
Jeongja stream 2.25 1.29 2.38 2.58 0.068 0.026 
Juja stream 2.03 1.17 1.98 1.62 0.034 0.015 
sum 6.13 8.97 10.53 11.45 0.227 0.446 
% difference 31.66 – 8.08 – 48.948 – 
Figure 6

Water quality simulation results based on the WQAM (Yongdam watershed).

Figure 6

Water quality simulation results based on the WQAM (Yongdam watershed).

Comparing the WQAM to the observed data in the BOD simulation, the percentage difference between the results of the WQAM and the observed water quality is 31.7%, which is in the ‘fair’ range according to the confidence range of the percentage difference of water quality (Donigian 2002). In the T-N simulation, the percentage difference is 8.1%, which is in the ‘very good’ range. The T-P simulation results did not match the observed data well. The percentage difference is 48.9%, which is out of the confidence range (Donigian 2002).

DISCUSSION

Parameters that influence water quality

In this research, imperviousness and land use are the main parameters used to develop the WQAM. Scenarios 1 to 5 and Scenarios 7 to 10 are based on imperviousness and land usage, respectively. Figure 7 shows that land usage exerts more of an influence on the simulated water quality than imperviousness. The BOD is influenced more by land usage (70%, S8, S9, S10) than by imperviousness (30%, S1, S2). T-N is influenced more by land usage (60%, S8, S9, S10) than by imperviousness (30%, S1, S2). In addition, T-P is also influenced more by land usage (80%, S8, S9, S10) than by imperviousness (20%, S1, S7).

Figure 7

The percentage of each selected scenario for the best WQAM based on non-linear regression. S1: Scenario 1, impervious; S2: Scenario 2, Impervious + pervious; S3: Scenario 3, Impervious + rainfall; S4: Scenario 4, Impervious + rainfall + slope; S5: Scenario 5, Impervious + slope; S6: Scenario 6, slope; S7: Scenario 7, Land use; S8: Scenario 8, Land use + rainfall; S9: Scenario 9, Land use + slope; S10: Scenario 10, Land use + rainfall + slope.

Figure 7

The percentage of each selected scenario for the best WQAM based on non-linear regression. S1: Scenario 1, impervious; S2: Scenario 2, Impervious + pervious; S3: Scenario 3, Impervious + rainfall; S4: Scenario 4, Impervious + rainfall + slope; S5: Scenario 5, Impervious + slope; S6: Scenario 6, slope; S7: Scenario 7, Land use; S8: Scenario 8, Land use + rainfall; S9: Scenario 9, Land use + slope; S10: Scenario 10, Land use + rainfall + slope.

Based upon the result of the WQAM selected from a non-linear regression method, we can assume that the water quality factors of BOD, T-N and T-P can be affected by land usage more than by imperviousness at this research site.

CONCLUSION

In this study, a WQAM is established through analysis of three cases and 10 scenarios, as shown in Table 5 by means of non-linear regression. It serves as an alternative to physical watershed modeling and was verified based on Yongdam Dam watershed as a case study. The data sets used as inputs to the equations included a pervious area, an impervious area, rainfall, slope, and land usage. Water quality results simulated through the WQAM produced are correlated with the observed data. The percentage difference of Yongdam Dam watershed BOD is 31.66%, T-N is 8.1%, and T-P is 48.9%.

Based on the percentage difference of the results, BOD and T-N meet the requirements of fair and very good ranges, respectively, according to the confidence range established by Donigian in 2002. However, the percentage difference result of the T-P simulation is below the fair range. Therefore, in order to establish a more accurate WQAM for a T-P simulation, we need to apply other statistics or methods in future research.

In addition, the WQAM is verified using the Yongdam Dam watershed as a case study. The WQAM shown in Table 7 can be adapted to simulate water quality in the Geum-Sum-Youngsan River watershed. The other watersheds, such as Han River watershed and Nakdong River watershed, should also be verified in order to select an appropriate WQAM for each watershed in future research.

Finally, the results show that the WQAM that relates the watershed parameters to watershed water quality, can be used as a rapid screening tool to aid determination of the best restoration locations.

REFERENCES

REFERENCES
Donigian
A. S.
2002
Watershed Model Calibration and Validation: The HSPF Experience, National TMDL Science and Policy 2002
, pp.
44
73
(
30)
.
Kutner
M. H.
Nachtsheim
C. J.
Neter
J.
Li
W.
2004
Applied Linear Statistical Models
.
Mehaffey
M. H.
Nash
M. S.
Wade
T. G.
Ebert
D. W.
Jones
K. B.
Rager
A.
2005
Linking land cover and water quality in New York City's water supply watershed
.
Environmental Monitoring and Assessment
107
,
29
44
.
Novotny
V.
2003
Water Quality Diffuse Pollution and Watershed Management. John Wiley & Sons, New York, USA, p 1–24
.
Novotny
V.
2008
Watershed Models, Encyclopedia of Ecology
, pp.
3748
3759
.
Pruden
A.
Mazdak Arabi
M.
Storteboom
H. N.
2012
Correlation between upstream human activities and riverine antiobiotic restistance genes
.
Environmental Science and Technology
46
,
11541
11549
.
Randhir
T. O.
O'Connor
R.
Penner
P. R.
Goodwin
D. W.
2001
A watershed-based land prioritization model for water supply protection
.
Forest Ecology and Management
143
(
2001
),
47
56
.
Schueler
T.
1987
Controlling Urban Runoff: A Practical Manual for Planning and Designing Urban Best Management Practices
.
Metropolitan Washington Council of Governments
.
Washington, DC, USA
.
Schueler
T.
2005
Urban Subwatershed Restoration Manual No. 1, An Integrated Framework to Restore Small Urban Watersheds
.
Center for Watershed Protection
,
Ellicott City, MD, USA
.
Singh
V. P.
Frevert
D. K.
2004
Watershed Modeling
.
World Water Congress
.
Wang
X.
Sheng
Yu
Huang
G. H.
2004
Land allocation based on integrated GIS-optimization modeling at a watershed level
,
Landscape and Urban Planning
66
(
2004
)
61
74
.
Woli
K. P.
Nagumo
T.
Kuramochi
K.
Hatano
R.
2004
Evaluating river water quality through land use analysis and N budget approaches in livestock farming areas
.
Science of the Total Environment
329
,
61
74
.

Supplementary data